© 2000 Oxford University Press
Nucleic Acids Research, 2000, Vol. 28, No. 21 4105–4112
A family of developmentally excised DNA elements in
Tetrahymena is under selective pressure to maintain an
open reading frame encoding an integrase-like protein
Jill A. Gershan and Kathleen M. Karrer*
Department of Biology, Marquette University, Milwaukee, WI 53201-1881, USA
Received July 28, 2000; Revised and Accepted September 8, 2000
ABSTRACT
Tlr1 is a member of a family of ∼20–30 DNA elements
that undergo developmentally regulated excision
during formation of the macronucleus in the ciliated
protozoan Tetrahymena. Analysis of sequence
internal to the right boundary of Tlr1 revealed the
presence of a 2 kb open reading frame (ORF)
encoding a deduced protein with similarity to retrotransposon integrases. The ORFs of five unique clones
were sequenced. The ORFs have 98% sequence
conservation and align without frameshifts, although
one has an additional trinucleotide at codon 561.
Nucleotide changes among the five clones are highly
non-random with respect to the position in the codon
and 93% of the nucleotide changes among the five
clones encode identical or similar amino acids,
suggesting that the ORF has evolved under selective
pressure to preserve a functional protein. Nineteen T/C
transitions in T/CAA and T/CAG codons suggest
selection has occurred in the context of the Tetrahymena genome, where TAA and TAG encode Gln.
Similarities between the ORF and those encoding
retrotransposon integrases suggest that the Tlr
family of elements may encode a polynucleotide
transferase. Possible roles for the protein in transposition of the elements within the micronuclear
genome and/or their developmentally regulated
excision from the macronucleus are discussed.
INTRODUCTION
DNA is remarkably mobile. Two major categories of DNA
rearrangements are those resulting from the invasion and
proliferation of independent transposable elements and those
which occur as part of a developmental program. The relationship
between the two kinds of events is not yet understood.
Transposons are invasive elements that integrate into host
genomes. Class I elements, or retrotransposons, transpose via
an RNA intermediate. Class II elements transpose via a DNA
DDBJ/EMBL/GenBank accession nos AF232242–AF232247
intermediate. Both types of elements encode polynucleotidyltransferases, which catalyze transposition, and an integrase or
a transposase in the class I and class II elements, respectively
(1). The active sites of these proteins are structurally similar in
that they contain regularly spaced acidic amino acids, called
DDE motifs, which chelate metal ions essential for enzymatic
activity.
In a variety of phylogenetically diverse organisms an integral
part of cellular development involves genomic reorganization
associated with developmentally regulated DNA deletion. One
example is the rearrangement of immunoglobulin genes and
T cell receptors of vertebrate immune systems. Recent
evidence has shown that the catalytic activity of RAG1, a
recombinase that mediates the rearrangement of these genes, is
dependent on three acidic residues, two of which may be
involved in binding metal ions (2,3). This raises the possibility
that there may be mechanistic similarities between the developmentally regulated DNA rearrangement and transposition.
Ciliated protozoa are particularly good model systems for
the analysis of DNA rearrangement because thousands of
programmed DNA deletions occur during development of the
somatic genome. Ciliates are single cell eukaryotes that
contain two structurally and functionally distinct nuclei that
share the same genetic origin. The transcriptionally active
somatic macronucleus sustains the cell during vegetative growth
while the germline micronucleus remains transcriptionally
silent. When ciliates reproduce sexually, the existing macronucleus is degraded and a new macronucleus develops from a
mitotic copy of the fertilization micronucleus. Macronuclear
development is associated with removal of thousands of
internal eliminated sequences (IESs) (4,5).
The developmentally regulated deletion of IESs has been
studied extensively in two classes of ciliates. In Oxytricha and
Euplotes (formerly hypotrichs, these ciliates are currently
placed in the class Spirotrichea; 6) ∼93% of the DNA is
eliminated during the transition from the micronuclear to the
macronuclear genome (5,7). Specific elements are removed
precisely and with 100% efficiency.
A substantial fraction of the eliminated DNA in the spirotrichs
belongs to families of thousands of repetitive transposon-like
elements. Examples of these include the Tec elements (transposon-like, Euplotes crassus) in Euplotes and the TBEs
*To whom correspondence should be addressed. Tel: +1 414 288 1474; Fax: +1 414 288 7357; Email: [email protected]
Present address:
Jill A. Gershan, Platelet Immunology, Blood Research Institute of Southeastern Wisconsin, 8727 Watertown Plank Road, Wauwatosa, WI 53226, USA
4106 Nucleic Acids Research, 2000, Vol. 28, No. 21
(telomere-bearing elements) of Oxytricha. Tec elements are
5.3 kb in length and have ∼700 bp of terminal inverted repeat
sequence (8). TBE elements are 4.1 kb in length and have
78 bp terminal inverted repeats which include 17 bp of
telomere-like G4T4 repeats (9–11). The Tec and TBE IESs are
thought to be class II transposable elements. They contain
terminal inverted repeats; there are short direct repeats that
resemble target site duplications at the ends of the elements
(5,10,12); and the presence of TBE1 is allele-specific, which is
consistent with a recent transposition event (13). Furthermore,
both types of elements have open reading frames encoding
putative transposases that are under selective pressure to
maintain a functional protein (11,14,15).
In addition to the transposon-like IESs, the micronuclear
genomes of the spirotrichs contain thousands of short, 10–539 bp,
non-coding IESs that are AT-rich. These IESs are unique or of
low copy number. One structural feature the short IESs of
Euplotes share with Tec elements is the presence of a terminal
TA direct repeat (5). Deletion of the two types of elements is
similar in that both are excised as circular molecules with
heteroduplex junctions (12) and their removal is coordinated
with DNA replication during polytenization of the DNA in the
macronuclear anlagen (16). Thus it seems likely that the two
types of elements are deleted by similar mechanisms.
The process of developmentally regulated DNA deletion
differs between spirotrichs and the evolutionarily distant
Oligohymenophora class of ciliates. The most extensively
studied member of the Oligohymenophora is Tetrahymena thermophila, wherein approximately 6000 elements, constituting 15%
of the genome, are deleted during macronuclear development.
As in the spirotrichs, IES deletion in Tetrahymena is 100%
efficient. However, in Tetrahymena excision is imprecise.
Several elements have alternative excision boundaries ranging
over a few hundred base pairs (17–20) and most show
sequence microheterogeneity at the rearrangement junctions
(20–22). Another striking difference is in the structure of the
excision products. In Tetrahymena the short IESs are more
likely to be excised as linear molecules, in contrast to the
circular heteroduplex excision products found in Euplotes
crassus (12,23).
Most of the IESs that have been analyzed to date in Tetrahymena are relatively short, ranging from 0.6 to 2.9 kb, and
consist of presumably non-coding AT-rich sequence (24). The
elements have no common structural features. Some have short
terminal direct repeats, but these are not required for excision
of the element (25). The largest deletion element characterized
to date is Tlr1, for Tetrahymena long repeat 1, which deletes
>25 kb of sequence. Tlr1 belongs to a family of ∼20–30 micronuclear-specific elements (19). The most striking structural
characteristic of Tlr1 is an 825 bp terminal inverted repeat near
the element termini. In this respect, Tlr1 structurally resembles
the transposon-like elements found in the spirotrichs.
In order to determine whether the Tlr elements contain an
open reading frame encoding a transposase, genomic clones of
several family members were isolated and sequenced. This
study describes the identification and sequence analysis of a
2 kb open reading frame within the Tlr family of elements.
This open reading frame is highly conserved and encodes a
putative protein with similarity to the integrases encoded by
retroviruses and retrotransposons.
MATERIALS AND METHODS
Accession numbers
The Tlr Int clone sequences have been deposited in the
GenBank database. Accession numbers are: Tlr1 Int,
AF232246; Tlr Int A, AF232242; Tlr Int B, AF232243; Tlr Int
C, AF232244; Tlr Int D, AF232245; Tlr Int E, AF232247.
Strains
Tetrahymena thermophila strain CU428 Mpr/Mpr (6-methylpurine-sensitive, VI) was obtained from Peter Bruns (Cornell
University, Ithaca, NY). Strain CU399 was used as a source for
the pMBR micronuclear plasmid library.
Micronuclear and macronuclear DNA isolation
Strain CU428 DNA was isolated according to the method
described by Gorovsky et al. (26). Following separation by
differential centrifugation, the nuclei were treated with 500 µg
proteinase K in 20 mM Tris–HCl pH 7.5, 40 mM EDTA and
1% SDS at 37°C for 1 h. DNA was extracted with phenol:chloroform (1:1) and ethanol precipitated.
Plasmid DNA isolation
Plasmid DNA was isolated from bacterial cell transformants
using either the Wizard Plus Maxiprep DNA purification protocol
(Promega) or the Qiagen plasmid purification protocol (Qiagen).
Oligonucleotides and primers
Oligonucleotides for circularization of HindIII–EcoRI fragments
of micronuclear DNA were 5′-AGCTTGAGCTCTCGAGTCGACGATCG-3′ and 5′-AATTCGATCGTCGACTCGAGAGCTCA-3′. The circularized DNA was amplified by inverse
PCR with primers 1 (5′-TCTATTCATCACTTTCTTA-3′) and
2 (5′-TTAATTTTATGTAAGTGAAGCTT-3′). Primer 3, (5′TCGATTTAAAATTATCTTTCTCTG-3′), derived from the
sequence of the inverse PCR product, was used against primer
2 to amplify Tlr family sequences from micronuclear DNA.
Tlr1 sequences were amplified with the Tlr1 specific primer 4
(5′-AATGTGAATTTCGATTCGAT-3′) and internal primer 5
(5′-GATGTCTACAATTTTATAGTTTCTC-3′).
Southern hybridization
Purified micronuclear DNA isolated from strain CU428 was
digested with the appropriate restriction endonucleases,
fractionated through 0.8–1% agarose gels and transferred via
capillary action onto NEF 976 Genescreen Plus transfer
membrane (NEN Research Products). DNA hybridization
probes were radioactively labeled by the random primer
protocol (Boehringer Mannheim). The nylon membranes were
washed in 2× SSC (1× SSC is 0.15 M NaCl, 0.015 M sodium
citrate) at room temperature, 2× SSC, 1% SDS at 65°C and
0.1× SSC at 58°C.
Genomic library screening
The pMBR plasmid library was constructed from strain CU399
micronuclear DNA partially digested with MboI and cloned
into the plasmid vector pUC19, as described previously (27).
Escherichia coli SURE (Promega) cells were transformed by
electroporation with 100 ng of the library. Cells were spread
onto Luria–Bertani broth (LB) plates containing 50 µg/ml
Nucleic Acids Research, 2000, Vol. 28, No. 21 4107
Figure 1. Restriction maps of the Tlr1 regions of micronuclear and macronuclear
DNA. The cross-hatched line represents macronucleus-destined sequence.
The thick solid line represents the micronuclear-limited sequence. The bold
arrows represent the 825 bp Tlr1 inverted repeat. The symbol * marks the location
of 19mer tandem repeats. The angled arrow depicts the location of the 2 kb
open reading frame identified in this study. Arrowheads represent PCR primers.
Restriction sites: B, BglII; C, ClaI; E, EcoRI; H, HindIII; S, Sau3A; X, XbaI.
The HindIII restriction site indicated by the dashed line is not present in Tlr1,
but is present in other family members. The boxes represent fragments used in
hybridization for isolation of Tlr Int clones.
carbenicillin. Colonies were screened by colony hybridization
according to the Southern hybridization protocol.
DNA sequencing
DNA was sequenced on an Applied Biosystems sequencer
using the ABI Prism™ dye terminator cycle sequencing ready
reaction kit with AmpliTaq DNA polymerase at the University
of Wisconsin, Milwaukee, automated sequencing center.
Sequence alignment
Sequences were aligned with Clustal 1.7 found under the
multiple sequence alignment program using the Baylor College
of Medicine search launcher (www.hgsc.bcm.tmc.edu/searchlauncher/ ) sequence utilities.
RESULTS
Cloning of the Tlr family of deleted elements
The Tlr1 element, with long inverted repeats near the termini,
is structurally similar to class II transposable elements, which
undergo ‘cut and paste’ transposition, often mediated by
transposases. The family of sequences is micronucleuslimited. The elements are unusual in that the copy number
differs for different regions of the inverted repeat. Southern
blot analysis showed that the innermost half of the Tlr1
inverted repeat hybridized to a family of 20–30 elements. This
defines the Tlr family. The outermost region of the Tlr1
inverted repeat has sequence homology to a subfamily of
sequences with a copy number of only 6–7 (19).
In order to determine whether the Tlr elements contained
open reading frames that might encode a transposase,
sequences internal to the inverted repeat at the right end of Tlr1
were cloned by inverse PCR (Fig. 1). A HindIII site is located
45 bp outside the right boundary of Tlr1. Genomic restriction
mapping placed an EcoRI site ∼2.5 kb inside the element.
Micronuclear DNA was digested with HindIII and EcoRI and
circularized by ligation of the DNA in the presence of two
Figure 2. Alignment of Tlr clones according to shared sequence. The vertical
line in clone Tlr1 Int indicates the location of the right Tlr1 boundary and the arrow
the inverted repeat. Mic, micronuclear-limited sequence; mac, macronucleusdestined sequence; angled arrow, the 2 kb open reading frame; triangles, the
primers used to PCR amplify clone Tlr1 Int from micronuclear DNA.
complementary oligonucleotides which hybridize to each other
to form a linker with HindIII and EcoRI sticky ends (Materials
and Methods). DNA sequences internal to the Tlr1 inverted
repeat were amplified by PCR with primers 1 and 2. The
resulting inverse PCR product was partially sequenced and
primer 3 was synthesized from this sequence. Micronuclear
DNA was PCR amplified using primers 2 and 3. Since both
primers are within sequences that are conserved in the Tlr
family, this reaction contained a mixture of PCR products
corresponding to various members of the family. A 1.2 kb
cloned fragment contained the expected sequences with
homology to the Tlr1 inverted repeat and 1079 bp of additional
sequence internal to the repeat.
The innermost 369 bp of the 1.2 kb PCR product contained
an open reading frame with similarity to Caenorhabditis
elegans retrotransposon CER1 integrase. This open reading
frame contains two Asp residues that are spaced proportionally
to the Asp residues in the D39–58D35E signature motif that is
present in all integrases. In order to obtain the complete open
reading frame, genomic clones were isolated from a plasmid
library of micronuclear DNA (Materials and Methods). The
library was screened with the 264 bp XbaI–HindIII and 971 bp
XbaI–XbaI fragments internal to the Tlr1 inverted repeat
(Fig. 1). Four clones with large inserts were selected for
sequencing. All of these clones contained a 2 kb open reading
frame encoding deduced proteins with homology to retroelement integrases. In Figure 2 the open reading frames of
clones Tlr Int A, Tlr Int B, Tlr Int C and Tlr Int D are aligned.
The sizes of the micronuclear DNA fragments demonstrate
that they were independent clones and slight differences in the
nucleotide sequence between clones indicated that they were
four different members of the Tlr family of elements.
Tlr Int A–D are members of a family of elements that is
repeated in the micronuclear genome. In order to obtain the
integrase-like ORF associated with Tlr1, the corresponding
fragment of micronuclear DNA was PCR amplified with Tlr1
specific primer 4, located in unique macronucleus-destined
DNA to the right of Tlr1, and primer 5, in conserved DNA 3′
to the open reading frame of the genomic clones (Fig. 2). The
sequence at the right end of the 4.4 kb PCR product matched
the known sequence of Tlr1 and the clone was designated Tlr1
Int. The integrase-like open reading frame begins 1583 nt from
4108 Nucleic Acids Research, 2000, Vol. 28, No. 21
Figure 3. Alignment of the first 227 amino acid residues in the Tlr Int consensus sequence with proteins or putative proteins that have similarity according to NCBI
BLAST. Amino acids that are identical in at least five sequences are shaded black. Amino acids that are similar in at least five sequences are shaded gray. The active
site DDE residues are marked with an asterisk. The integrase HHCC domain residues are indicated by circles. E values and accession numbers of the sequences
are: C.elegans cosmid Y57G11C, 1e–22, Z99281; X.maculatus HASI putative protein, 3e–22, U43331; Trichopulsia ni TED retrotransposon, 7e–11, B36329;
Drosophila melanogaster mdg3 retrotransposon, 1e–7, X95908; Schizosaccharomyces pombe Tf1 retrotransposon 1e–7, M38526; Zea mays Opie-2 retrotransposon,
2e–7, AF105716; S.cerevisiae Ty1 retrotransposon, 4e–5, Z47746.
the right Tlr1 boundary and 710 nt internal to the inverted
repeat (Fig. 1).
Similarity of the conceptual protein to retroelement
integrases
The Tlr1 Int clone and Tlr Int A–C each contained an open
reading frame of 1998 nt encoding a conceptual protein of 666
amino acids. The open reading frame of Tlr Int D encoded one
additional amino acid, due to insertion of a trinucleotide
encoding Glu at residue 561. To find similarity to previously
identified proteins, the DNA sequence of the 2 kb open reading
frame was submitted to NCBI BLASTX (Basic Local Alignment
Search Tool) (28). The closest matches to the Tlr family open
reading frame were sequences found in transposons or putative
transposons. An alignment of the deduced proteins from a
variety of organisms is shown in Figure 3. The best characterized
of these was the TYB protein of Saccharomyces cerevisiae,
which encodes a retrotransposon integrase. All integrases and
many transposases contain three acidic amino acids with
characteristic spacing (D39–58D35E) which comprise the catalytic
core of the proteins. Each of these residues is independently
essential for catalytic activity involved in the reactions of
transposition (29–35). The Tetrahymena open reading frame
encodes these acidic residues, along with characteristic blocks
of conserved amino acid residues surrounding each of the
signature amino acids.
Integrases and transposases can be distinguished on the basis
of the conserved amino acids surrounding the DDE signature
(36; Fig. 4). The first Asp of the signature is generally followed
by an acidic residue in transposases and more often by a hydrophobic amino acid in integrases. The second Asp is followed
by an Asn in both classes of enzymes. Most notably, the
integrase Glu is in a consensus sequence ERMNR/KTI/LK. In
the transposases the consensus sequence in the region of the
Glu is SPDLNPIEHL/I. The alignment in Figure 4 suggests
that the deduced protein encoded by the Tlr family of elements
in Tetrahymena is more closely related to the integrases than
Nucleic Acids Research, 2000, Vol. 28, No. 21 4109
42 encoded similar amino acids and 14 encoded dissimilar
amino acids (Table 1). Therefore, 93% of the nucleotide
changes altered codons to encode either identical or similar
amino acids and the amino acid sequence of the putative
proteins encoded by the five elements is 99% conserved.
Table 1. Nucleotide and amino acid substitutions among the Int open reading
frames
Amino acid
Figure 4. An alignment of the residues surrounding the DDE amino acids in
integrases and transposases with the corresponding Tlr1 Int amino acids. The
amino acids common to all types of proteins are red. Blue letters correspond to the
retroviral integrase consensus, green letters correspond to the retrotransposon
integrase consensus and purple letters correspond to the class II element
transposase consensus. Black letters show no consensus. Similar residues are
included in the consensus and similarity for this analysis is defined as a score
of 5 on the SG Matrix scoring system (38).
transposases. Thus this sequence is referred to as the Int open
reading frame.
In addition to the DDE signature motif, integrases have an
N-terminal zinc finger domain (HHCC) (Fig. 3), which is not
found in transposases. This domain is required for the formation
of a stable complex between viral cDNA and the integrase
during cDNA 3′-end processing and strand transfer (30,37).
Interestingly, although the DDE signature of the Tlr Int open
reading frame is more similar to integrases than transposases,
it does not have this zinc finger domain. The uncharacterized
elements in C.elegans and Xiphophorus maculatus that are
most similar to the Tlr Int open reading frame also lack this
HHCC motif (Fig. 3).
Conservation of the open reading frame
The open reading frames encoding the integrase-like proteins
of Tlr1 and Tlr Int A–D were analyzed to determine the degree
of conservation of the putative gene products. There were 196
nucleotide changes among the 9993 nt analyzed, for 98%
conservation of sequence. According to χ2 analysis, the
position of the nucleotide changes within the open reading
frames was highly non-random (P << 0.001), with most of
them occurring in the third position of the codon where there is
the greatest likelihood of amino acid conservation (Table 1).
To assess the degree of amino acid conservation, amino acid
substitutions were categorized as identical, similar or dissimilar.
Similarity was defined according to the Structure–Genetic
(SG) Matrix scoring system (38), where amino acids are
assigned a number from 0 to 6 based on the structural identity
and likelihood of interchanges. A score of 6 indicates identity.
For the analysis in Table 1, similarity was defined as an SG
score of 4 or 5. Of the 196 nucleotide changes among the five
open reading frames, 140 encoded identical amino acids,
Nucleotide position
Total nucleotides
First
Second
Third
Identical
20
0
120
140 (72%)
Similar
13ab
16
13ab
42 (21%)
Dissimilar
4c
9cd
1d
14 (7%)
Total
37
25
134
196
a–dFour
codons with two nucleotides different from the consensus sequence.
Conservation of the amino acid sequence despite nucleotide
changes suggested that the open reading frame has evolved
under selective pressure to conserve a functional protein.
Statistical analysis of the synonymous versus non-synonymous
nucleotide changes was done in order to determine the significance of the amino acid conservation (39). According to this
method, nucleotide substitutions that encode similar or identical amino acids are synonymous (ps) whereas nucleotide
substitutions that encode dissimilar amino acids are nonsynonymous (pn). If there is selective pressure to conserve a
protein coding region, ps is expected to be greater than pn. In a
one-tailed t-test the ps value was significantly greater than pn
for each pair of open reading frames at P << 0.001.
One unusual feature of the Tetrahymena genetic code is that
the canonical stop codons TAA and TAG encode Gln (40,41).
Thus T→C transitions in the first nucleotide of T/CAA or T/CAG
Gln codons are silent. Such transitions accounted for 19 of the
20 nucleotide changes in the first nucleotide of a codon that
resulted in identical amino acids (Table 1), suggesting that the
open reading frame has evolved in the context of the Tetrahymena genome.
The N-terminus of the Tetrahymena integrase-like protein
had the highest degree of similarity to integrases of other
organisms. In order to determine whether selection among the
Tlr family members is uniform over the length of the putative
protein, the open reading frame was arbitrarily divided
into thirds, with the DDE motifs of the active site located in the
N-terminal third of the protein. Interestingly, χ2 analysis
indicated that although the nucleotide substitutions occur
randomly across the gene, dissimilar amino acid changes were
preferentially located in the C-terminal third of the open
reading frame (P < 0.01). Thus, although the C-terminal region
of the protein is under strong selection, it is not as stringent as
the selection in the N-terminal two-thirds of the protein.
Copy number and genome specificity of the Tlr family of
sequences
Whereas the innermost portion of the Tlr1 inverted repeat is
repeated 20- to 30-fold in the micronuclear genome, the outermost sequences, as indicated by hybridization of the tandemly
4110 Nucleic Acids Research, 2000, Vol. 28, No. 21
Figure 5. (A) Partial restriction map of restriction sites in the Tlr Int clones.
Angled arrow, Int open reading frame; heavy arrow, Tlr1 inverted repeat; bars,
the HindIII–XbaI and EcoRV–ClaI fragments used to probe the blots in (B)
and (C), respectively. Restriction sites C, ClaI; H, HindIII; V, EcoRV; X, XbaI.
(B) Southern blot of micronuclear (MI) and macronuclear (MA) DNA digested
with EcoRI, XbaI, EcoRV and HindIII. (C) Southern blot of micronuclear (MI)
and macronuclear (MA) DNA digested with ClaI + EcoRV, EcoRV, HindIII
and RsaI.
during macronuclear development. The presence of macronuclear DNA was confirmed by ethidium bromide staining of
the gel prior to transfer (data not shown).
A similar copy number was found for sequences in the open
reading frame. Sequence analysis of the five clones at hand
revealed a conserved EcoRV site within the open reading
frame and no additional EcoRV sites in the conserved
sequence 5′ to the open reading frame. Therefore, Southern
analysis of DNA digested with EcoRV and probed with the
ClaI–EcoRV fragment located entirely within the open reading
frame provided an estimate of the number of elements that
contain the open reading frame (Fig. 5C). The observed pattern
of 20–30 bands was similar in number to that found for
sequences 5′ to the open reading frame and for the innermost
part of the Tlr1 inverted repeat. As in the previous Southern
blot, the probe hybridized only to micronuclear DNA,
suggesting that the 2 kb open reading frame is located within a
family of micronuclear-limited elements.
The similarity in copy number and the micronucleus-limited
character suggested that the integrase-like open reading frame
may be generally associated with sequences in the Tlr1
inverted repeat. The limited data available for individual
clones supports this model. Of the clones that cover the
integrase-like open reading frame, only Tlr Int A extends far
enough 5′ of the gene to contain DNA with homology to the
Tlr1 inverted repeat. Another clone, Tlr Int E, which contained
only the 5′-end of the integrase-like open reading frame,
extended far enough to include sequences with homology to
the Tlr1 inverted repeat. Sequence data from these three clones
suggests that the integrase-like open reading frame is generally
associated with the Tlr family, among which the innermost
∼275 bp of the Tlr1 inverted repeat are conserved.
The Southern blots also support the sequence data indicating
a high degree of conservation among the integrase-like genes.
Three major bands were seen in DNA digested with ClaI and
EcoRI (Fig. 5C). One of these had the mobility expected for
the 937 bp ClaI–EcoRV fragment that is present in all of the
sequenced clones. The two other bands at 1.4 and 3.5 kb
indicate that there are three major variations in the ClaI +
EcoRV restriction pattern among the various family members.
DISCUSSION
repeated 19mers, belongs to a smaller family of only six to
seven elements (19). Thus it was of interest to determine the
copy number of sequences internal to the inverted repeat and of
the open reading frame encoding the putative integrase. Micronuclear and macronuclear DNA were digested with various
restriction enzymes and hybridized in a Southern analysis with
the 714 bp HindIII–XbaI fragment containing sequences
between the Tlr1 inverted repeat and the integrase open
reading frame (Fig. 5B). In the lanes containing micronuclear
DNA digested with EcoRI, EcoRV and HindIII ∼20–30 fragments
hybridized, suggesting that sequences homologous to this
fragment have a copy number similar to that of the inner part of
the Tlr1 inverted repeat. In micronuclear DNA digested with
XbaI the majority of the hybridization was to a 1 kb fragment.
This was consistent with the sequence data suggesting that
sequences within the element are highly conserved among
family members. There was no hybridization to macronuclear
DNA, thus the family of elements is efficiently eliminated
The Tlr family of micronucleus-limited elements in Tetrahymena
encodes a putative protein with similarity to retrotransposon
integrases. The open reading frames from five of the 20–30
family members have been cloned and sequenced. Nucleotide
changes among the five family members are highly nonrandom, suggesting that the elements are under selective
pressure to maintain a functional protein in most or all of the
elements.
The role of the integrase-like protein in Tetrahymena
biology is unknown. One possibility is that the Tlr family of
IESs encodes the machinery for its own developmentally
regulated excision and the integrase-like enzyme is a polynucleotidyltransferase that is part of that machinery. A functional
relationship between transposable elements and developmentally
regulated DNA excision was first proposed to account for
conservative selection of the open reading frames in the Tec
and TBE elements of the spirotrichs (5). According to this
model, IESs are degenerate transposons. Developmentally
Nucleic Acids Research, 2000, Vol. 28, No. 21 4111
regulated DNA deletion evolved as a mechanism to remove
them from the transcriptionally active macronucleus, with the
transposase serving as the excisase. In Euplotes and Oxytricha,
where many of the IESs occur within coding sequences,
efficient IES excision might be expected to exert a powerful
selective pressure to conserve the active transposase (42). This
model is somewhat less compelling in Tetrahymena, where no
protein coding sequences are known to be interrupted by IESs.
However, it is possible that there are other selective advantages
to removing IESs from the macronucleus genome. For
example, they might play a role in micronuclear functions such
as chromatin condensation and/or transcriptional silencing that
are dispensable in the macronucleus (43).
Although conservative selection for functional protein might
be expected if the integrase-like open reading frame of the Tlr
family encodes excisase, it is not clear what selective mechanism
would maintain functional genes in multiple copies of the
element, since such an enzyme would presumably act in trans.
Constructs containing the inverted repeats of Tlr1, but lacking
the integrase-like gene, undergo efficient and accurate rearrangement in vivo (44). Thus developmentally regulated DNA excision
of the Tlr family does not require an active integrase-like gene
in cis.
Multiple genes encoding the integrase-like protein might be
required to synthesize sufficient amounts of the protein to
catalyze excision of the family of elements within a specific
and brief developmental time period. However, transcripts of
the Tec transposase gene of Euplotes were detected only by
extremely sensitive methods involving Southern hybridization
of RT–PCR products and the transcript levels are thought to be
insufficient to account for the en masse excision of ∼106 Tec
elements/polytene nucleus in a 2–4 h time period (45,46).
Similarly, transcripts of the integrase-like genes in Tetrahymena
have not been detected by standard northern blot analysis
(J.A.Gershan, unpublished data).
A second possible function of the integrase-like gene is that
the Tlr family of sequences are mobile elements and the
integrase-like enzyme is responsible for their transposition in
the micronuclear genome. Although the enzyme encoded by
the Tlr family of elements has a DDE motif similar to that of
retroelements, the Tlr elements differ from retrotransposons in
several respects. At >25 kb (J.D.Wuitschick and J.A.Gershan,
unpublished data), they are larger than most active retroelements and the terminal repeats are inverted rather than
direct repeats. Analysis of sequences surrounding the Int open
reading frame has not revealed a discernible gag–pol gene
structure characteristic of retroelements and the putative
integrase gene lacks the N-terminal HHCC zinc finger domain
that is required for retroelement cDNA processing (30,37). (A
complete analysis of the structure of these large elements is in
progress and will be published elsewhere.) The long inverted
repeats of Tlr1 are a structural characteristic common to many
class II elements. Perhaps the Tlr elements are class II
elements in which the motifs surrounding the active site DDE
residues of the transposases are more similar to those in the
integrases than to those of the majority of transposases. This
would be unusual, but not unprecedented (36). The bacterial IS30
element is an example of a class II element with integrase-like
active site motifs. The putative enzyme encoded by the Tlr
family is like the transposase of the bacterial IS30 elements in
that both lack the N-terminal HHCC integrase domain (47).
If the Tlr family of sequences are transposons, then it is
necessary to account for the fact that these elements are found
exclusively in the 15% of the micronuclear genome that undergoes developmentally regulated elimination. Excision of the
Tlr elements during macronuclear development could be a
secondary consequence of transposon targeting into micronuclear IESs. Two observations support this hypothesis. First,
Tlr1 is removed from the macronuclear genome at the same
stage of macronuclear development as the smaller, AT-rich
IESs (17; Capowski and K.M.Karrer, unpublished data).
Second, flanking sequences have been shown to play a role in
the delineation of the rearrangement boundaries of Tlr1 (44).
This is not an expected feature of transposon excision, but is
consistent with the hypothesis that Tlr1 resides within an IES
because, in Tetrahymena, cis-acting sequences in the flanking
DNA regulate the deletion of IESs (21,22,48).
Unique features of chromatin structure might serve to target
transposition of Tlr elements to IESs. There is a precedent for
transposon targeting to regions of distinct chromatin structure
in yeast, where Ty5 elements are selectively inserted into
regions of silent chromatin by association of the integration
complexes with localized host factors (49–52). Differences
between the chromatin structure of eliminated sequences and
macronucleus-destined sequences have been detected in
Euplotes by analysis of chromatin digested with micrococcal
nuclease (53). In Tetrahymena the chromatin of IES-containing
regions of the genome is distinct from that of the bulk of the
genome during macronuclear development. The abundant
stage-specific proteins Pdd1p and Pdd2p, both of which are
required for developmentally programmed DNA deletion in
Tetrahymena (54,55), are preferentially associated with
heterochromatic regions containing IESs (56,57).
If the integrase-like protein encoded by the Tlr family of
elements is a transposase and is not functioning as the excisase,
then the need to maintain sufficient enzyme for IES excision
cannot be invoked to explain the apparent selection on these
genes. The transposases of class II elements can be subject to
selective pressure for functions other than the transposase
activity, such as repression of transposition (58).
Five of the estimated 20–30 integrase-like open reading
frames, a significant fraction of the total Tlr family, were
analyzed in this study. Despite numerous nucleotide changes
amongst the genes, all of the family members examined maintained the open reading frame and a high degree of protein
similarity. Whatever the biological role of the integrase-like
protein, the data indicate that there is strong selective pressure
to maintain an active gene in most or all of the Tlr family
members.
ACKNOWLEDGEMENTS
We thank Inta Kalve and Kevin Miller for technical assistance.
This work was supported by grant MCB-9974885 from the
National Science Foundation. Jill Gershan was supported in
part by a Marquette University John P. Raynor fellowship and
an Arthur J. Schmitt Foundation fellowship.
REFERENCES
1. Polard,P. and Chandler,M. (1995) Mol. Microbiol., 15, 13–23.
4112 Nucleic Acids Research, 2000, Vol. 28, No. 21
2. Landree,M.A., Wibbenmeyer,J.A. and Roth,D.B. (1999) Genes Dev., 13,
3059–3069.
3. Kim,D.R., Dai,Y., Mundy,C.L., Yang,W. and Oettinger,M.A. (1999)
Genes Dev., 13, 3070–3080.
4. Karrer,K.M., (1999) In Asai,D.J. and Forney,J.D. (eds), Tetrahymena
thermophila. Academic Press, San Diego, CA, pp. 127–186.
5. Klobutcher,L.A. and Herrick,G. (1997) Prog. Nucleic Acid Res. Mol.
Biol., 56, 1–62.
6. Lynn,D.H. and Small,E.B. (1997) Rev. Soc. Mex. Hist. Nat., 47.
7. Swanton,M.T., Greslin,A.F. and Prescott,D.M. (1980) Chromosoma, 77,
203–215.
8. Jahn,C.L., Krikau,M.F. and Shyman,S. (1989) Cell, 59, 1009–1018.
9. Herrick,G., Cartinhour,S., Dawson,D., Ang,D., Sheets,R., Lee,A. and
Williams,K. (1985) Cell, 43, 759–768.
10. Hunter,D.J., Williams,K., Cartinhour,S. and Herrick,G. (1989)
Genes Dev., 3, 2101–2112.
11. Doak,T.G., Witherspoon,D.J., Doerder,F.P., Williams,K. and Herrick,G.
(1997) Genetica, 101, 75–86.
12. Jaraczewski,J.W. and Jahn,C.L. (1993) Genes Dev., 7, 95–105.
13. Seegmiller,A., Williams,K.R., Hammersmith,R.L., Doak,T.G.,
Witherspoon,D., Messick,T., Storjohann,L.L. and Herrick,G. (1996)
Mol. Biol. Evol., 13, 1351–1362.
14. Jahn,C.L., Doktor,S.Z., Frels,J.S., Jaraczewski,J.W. and Krikau,M.F.
(1993) Gene, 133, 71–78.
15. Doak,T.G., Doerder,F.P., Jahn,C.L. and Herrick,G.H. (1994) Proc. Natl
Acad. Sci. USA, 91, 942–946.
16. Frels,J.S. and Jahn,C.L. (1995) Mol. Cell. Biol., 15, 6488–6495.
17. Austerberry,C.F., Allis,C.D. and Yao,M.-C. (1984) Proc. Natl Acad. Sci.
USA, 81, 7383–7387.
18. Austerberry,C.F. and Yao,M.-C. (1988) Mol. Cell. Biol., 8, 3947–3950.
19. Wells,J.M., Ellingson,J.L.E., Catt,D.M., Berger,P.J. and Karrer,K.M.
(1994) Mol. Cell. Biol., 14, 5939–5949.
20. Patil,N.S., Hempen,P.M., Udani,R.A. and Karrer,K.M. (1997)
J. Eukaryot. Microbiol., 44, 518–522.
21. Austerberry,C.F., Snyder,R.O. and Yao,M.-C. (1989) Nucleic Acids Res.,
17, 7263–7272.
22. Li,J. and Pearlman,R.E. (1996) Nucleic Acids Res., 24, 1943–1949.
23. Yao,M.-C. and Yao,C.-H. (1994) Nucleic Acids Res., 22, 5702–5708.
24. Wuitschick,J.D. and Karrer,K.M. (1999) J. Eukaryot. Microbiol., 46,
239–247.
25. Godiska,R. and Yao,M.-C. (1990) Cell, 61, 1237–1246.
26. Gorovsky,M.A., Yao,M.-C., Keevert,J.B. and Pleger,G.L. (1975)
Methods Cell Biol., 9, 311–327.
27. Rogers,M.B. and Karrer,K.M. (1989) Dev. Biol., 131, 261–268.
28. Altschul,S.F., Madden,T.L., Schaffer,A.A., Zhang,Z., Miller,W. and
Lipman,D.J. (1997) Nucleic Acids Res., 25, 3389–3402.
29. Drelich,M., Wilhelm,R. and Mous,J. (1992) Virology, 188, 459–468.
30. Engelman,A. and Craigie,R. (1992) J. Virol., 66, 6361–6369.
31. Kulkosky,J., Jones,K.S., Katz,R.A., Mack,J.P.G. and Skalka,A.M. (1992)
Mol. Cell. Biol., 12, 2331–2338.
32. Leavitt,A.D., Shive,L. and Varmus,H.E. (1993) J. Biol. Chem., 268,
2113–2119.
33. Jenkins,T.M., Esposito,D., Engelman,A. and Craigie,R. (1997) EMBO J.,
16, 6849–6859.
34. Wiskerchen,M. and Muesing,M.A. (1995) J. Virol., 69, 376–386.
35. van Gent,D.C., Oude Groeneger,A.A.M. and Plasterk,R.H.A. (1992)
Proc. Natl Acad. Sci. USA, 89, 9598–9602.
36. Capy,P., Vitalis,R., Langin,T., Higuet,D. and Bazin,C. (1996) J. Mol. Evol.,
42, 359–368.
37. Ellison,V. and Brown,P.O. (1994) Proc. Natl Acad. Sci. USA, 91, 7316–7320.
38. Feng,D.F., Johnson,M.S. and Doolittle,R.F. (1985) J. Mol. Evol., 21, 112–125.
39. Nei,M. and Gojobori,T. (1986) Mol. Biol. Evol., 3, 418–426.
40. Horowitz,S. and Gorovsky,M.A. (1985) Proc. Natl Acad. Sci. USA, 82,
2452–2455.
41. Hanyu,N., Kuchino,Y. and Susumu,N. (1986) EMBO J., 5, 1307–1311.
42. Witherspoon,D.J., Doak,T.G., Williams,K.R., Seegmiller,A., Seger,J. and
Herrick,G. (1997) Mol. Biol. Evol., 14, 696–706.
43. Yao,M.-C. (1996) Trends Genet., 12, 26–30.
44. Patil,N. and Karrer,K.M. (2000) Nucleic Acids Res., 28, 1465–1472.
45. Klobutcher,L.A., Turner,L.R. and LaPlante,J. (1993) Genes Dev., 7, 84–94.
46. Jaraczewski,J.W., Frels,J.S. and Jahn,C.L. (1994) Nucleic Acids Res., 22,
4535–4542.
47. Dalrymple,B., Caspers,P. and Arber,W. (1984) EMBO J., 3, 2145–2149.
48. Chalker,D.L., La Terza,A., Wilson,A., Kroenke,C.D. and Yao,M.C.
(1999) Mol. Cell. Biol., 19, 5631–5641.
49. Zou,S., Ke,N., Kim,J.M. and Voytas,D. (1996) Genes Dev., 10, 634–645.
50. Zou,S. and Voytas,D.F. (1997) Proc. Natl Acad. Sci. USA, 94, 7412–7416.
51. Gai,X. and Voytas,D.F. (1998) Mol. Cell, 1, 1051–1055.
52. Zhu,Y., Zou,S., Wright,D.A. and Voytas,D.F. (1999) Genes Dev., 13,
2738–2749.
53. Jahn,C.L. (1999) Mol. Biol. Cell, 10, 4217–4230.
54. Coyne,R.S., Nidiforov,M., Smothers,J.F., Allis,C.D. and Yao,M.-C.
(1999) Mol. Cell, 4, 865–872.
55. Nikiforov,M.A., Smothers,J.F., Gorovsky,M.A. and Allis,C.D. (1999)
Genes Dev., 13, 2852–2862.
56. Madireddi,M.T., Coyne,R.S., Smothers,J.F., Mickey,K.M., Yao,M.-C.
and Allis,C.D. (1996) Cell, 87, 75–84.
57. Smothers,J.F., Mizzen,C.A., Tubbert,M.M., Cook,R.G. and Allis,C.D.
(1997) Development, 124, 4537–4545.
58. Misra,S. and Rio,D.C. (1990) Cell, 62, 269–284.